Ask a big language version (LLM) like GPT-4 to scent a rain-soaked campground, and it’ll pleasantly decrease. Ask the very same system to define that fragrance to you, and it’ll wax poetic regarding “an air close expectancy” and “an aroma that is both fresh and natural,” regardless of having neither previous experience with rainfall neither a nose to aid it make such monitorings. One feasible description for this sensation is that the LLM is just simulating the message existing in its huge training information, instead of collaborating with any type of genuine understanding of rainfall or odor.
发布者：Dr.Durant，转转请注明出处：https://robotalks.cn/experiments-reveal-llms-develop-their-own-understanding-of-reality-as-their-language-abilities-improve/
